The authors collected county-level data from the US Census on 3,142 counties and stratified them by the population of black Americans in the community (in relation to whether the community's black population was greater or less than the national average). The authors used Bayesian hierarchical models to control for sociodemographics, comorbidities, and socioeconomic determinants, including adjustments for county-specific characteristics (from Table 1) and temporality for the epidemic. Exponentiated regression coefficients were computed and rate ratios for COVID-19 morbidity and mortality were analyzed.